特大城市交通走廊通勤客流Nested-Logit模型研究  

Nested-Logit model of combined selection for commuters analysis in urban transportation corridor

作　　者：沈犁[1] 张桐[1] 王周全[1] SHEN Li;ZHANG Tong;WANG Zhouquan(School of Transportation and Logistics, Southwest Jiaotong University, Chengdu 610031, China)

机构地区：[1]西南交通大学交通运输与物流学院,成都610031

出　　处：《计算机工程与应用》2016年第21期229-234,共6页Computer Engineering and Applications

基　　金：国家自然科学基金(No.51108390)

摘　　要：为了科学分析特大城市交通客运走廊内高峰时段通勤者的出行选择特征,以及交通需求管理措施施行时不同时段、不同方式之间的客流流向规律,研究了走廊内高峰时段通勤者基于出行方式与出发时刻的二维联合选择问题,建立了方式选择枝位于上层而时刻选择枝位于下层的联合选择Nested-Logit模型。其后通过以成都市中心城南北向交通客运走廊为例的通勤者样本数据的分析,采用上述联合选择NL模型进行了高峰时段交通管控措施影响效果的模拟预测,以探究实现客运走廊不同方式、时段内交通流量均衡分布的有效策略与措施。研究结果表明,联合选择NL模型相比一般MNL模型更加合理且具有更优的统计特性,且能够真实反映特大城市交通客运走廊内高峰时段通勤客流的分布情况,模拟施行交通需求管理措施对通勤者的出行选择的影响,为交通客运走廊内各交通方式结构以及运输组织模式的优化提供理论依据与决策支持。To analyze the traveling selection characteristics of commuters in urban transportation corridor in morning peak period and the discipline of passenger flow when traffic management & control measures are implemented, this paper focuses on the combined selection problem about commuters’mode choice behavior and departure-time along the corridor, then builds a Nested-Logit model which gets departure time located in lower layer. By analyzing the sample data of commuter travel along the north-south corridor in Chengdu city to explore effective tactics and measures, the NL model is adopted to simulate the situation with traffic control measures implemented. The result shows that NL model is more rational and exhibits better statistic characteristic compared with traditional MNL model, and it reflects the real distribution of commuters at peak hours in urban transportation corridor, calculates the influence of traffic management & control measures on commuters’mode choice behavior, providing important support for optimizing the mode structure and improving service efficiency of urban transportation corridors.

关 键 词：城市交通走廊 联合选择 出行方式 出行时刻 NL模型 交通管控措施 

分 类 号：TP39[自动化与计算机技术—计算机应用技术]